Biochemistry 1981, 20, 6051-6060 Farese, R. V., Sabir, M. A,, & Larson, R. E. (1980~)Proc. Natl. Acad. Sci. U.S.A. 77, 7189. Farese, R. V., Larson, R. E., & Sabir, M. A. (198Od) Biochim. Biophys. Acta 633, 479. Farese, R. V., Bidot-Lopez, P., Sabir, M. A., Smith, J., Schinbeckler, B., & Larson, R. (1980e) J . Clin. Zmest. 65, 1523. Farese, R. V., Sabir, M. A., & Larson, R. E. (19800 J . Clin. Invest. 66, 1428. Ferguson, J. J., Jr. (1963) J . Biol. Chem. 238, 2754. Garren, L. E., Ney, R. L., & Davis, W. W. (1965) Proc. Natl. Acad. Sci. U.S.A. 53, 1443. Haynes, R. C., & Berthet, L. (1957) J . Biol. Chem. 225,115. Hokin, L. E., & Hokin, M. R. (1958) J . Biol. Chem. 233,805.
605 1
Hokin, M. R., & Hokin, L. E. (1959) J . Biol. Chem. 234, 1381. Hosaka, K., Yamashita, S., & Numa, S . (1975) J . Biochem. (Tokyo) 77, 501. Kabara, J. J., & Chen, J. S . (1976) Anal. Chem. 48, 814. Lamb, R. E., & Fallon, H. J. (1970) J. Biol. Chem. 254,3075. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., & Randall, J. (1951) J . Biol. Chem. 193, 265. Martensson, E., & Kanfer, J. (1968) J . Biol. Chem. 243,497. Merlie, J. P., & Pizer, L. I. (1973) J . Bacteriol. 116, 355. Michell, R. H. (1975) Biochim. Biophys. Acta 414, 81. Silber, R. H., Busch, R., & Oslapas, R. (1958) Clin. Chem. ( Winston-Salem, N.C.) 4, 278.
Divalent Metal Ion, Inorganic Phosphate, and Inorganic Phosphate Analogue Binding to Yeast Inorganic Pyrophosphatase+ Barry S . Cooperman,* Anna Panackal, Bleecker Springs, and Donald J. Hamm*
ABSTRACT:
Four different techniques, equilibrium dialysis, protection of enzymatic activity against chemical inactivation, 31Prelaxation rates, and water proton relaxation rates, are used to study divalent metal ion, inorganic phosphate, and inorganic phosphate analogue binding to yeast inorganic pyrophosphatase, EC 3.6.1.1. A major new finding is that the binding of a third divalent metal ion per subunit, which has elsewhere been implicated as being necessary for enzymatic activity [Springs, B., Welsh, K. M., & Cooperman, B. S . (1981) Biochemistry (in press)], only becomes evident in the presence of added inorganic phosphate and that, reciprocally, inorganic phosphate binding to both its high- and low-affinity sites on the enzyme is markedly enhanced in the presence of
divalent metal ions, with Mn2+ causing an especially large increase in affinity. The results obtained allow evaluation of all of the relevant equilibrium constants for the binding of Mn2+ and inorganic phosphate or of Coz+ and inorganic phosphate to the enzyme and show that the high-affinity site has greater specificity for inorganic phosphate than the lowaffinity site. In addition, they provide evidence against divalent metal ion inner sphere binding to phosphate for enzyme subunits having one or two divalent metal ions bound per subunit and evidence for a conformational change restricting active-site accessibility to solvent on the binding of a third divalent metal ion per subunit.
m r k in this laboratory over the last few years has been devoted to elucidating the mechanism of action of the dimeric enzyme yeast inorganic pyrophosphatase, EC 3.6.1.1 (PPase).' This enzyme catalyzes both pyrophosphate (PPI) hydrolysis and water-phosphate (Pi) oxygen exchange. In recent papers (Hamm & Cooperman, 1978; Springs et al., 1981) we have presented a unified scheme for PPase action which accounts quantitatively for the observed overall rate constants for both processes and, in addition, presented evidence that three Mg2+ per active subunit are required for enzymatic activity and that there are two Pi sites per subunit, one of high and one of low affinity. We will henceforth denote these sites as site 1 and site 2, respectively. Although Mg2+ confers maximal activity on PPase, appreciable activity is also found in the presence of ZnZ+,Co2+, and MnZ+(Butler & Sperow, 1977; Janson et al., 1979). In this paper we use four different techniques, equilibrium dialysis, protection of enzymatic activity against chemical
modification, 31Prelaxation rates, and water proton relaxation rates, to investigate divalent metal ion, Pi, and Pi analogue (methylphosphonate, thiophosphate, and phosphoramidate) binding to PPase. The binding of Pi in the presence of either Mn2+or Co2+has been investigated in the most detail, because the paramagnetism of these two metal ions will make them extremely useful in future N M R and ESR studies. Our major findings and conclusions are as follows. (1) In addition to the two divalent metal ions bound with approximately equal affinity per PPase subunit in the absence of Pi, a third divalent metal ion is bound with approximately equal affinity in the presence of Pi. (2) Reciprocally, the affinity of Pi for both of its sites on PPase is markedly increased in the presence of divalent metal ion. In the presence of either Mn2+or Co2+,the binding to site 1 is strong enough to permit direct stoichiometric titration. (3) The equilibrium dialysis and 31Prelaxation rate studies permit evaluation of all of the relevant equilibrium constants for the binding of three MnZ+
t From the Department of Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania 19104. Received April 16,1981. This work was supported by a research grant from the National Institutes of Health (AM- 13212). 'Present address: Best Food, a unit of CPC International, Union, NJ 07083.
Abbreviations used: BSA, bovine serum albumin; PEI, poly(ethy1enimine); EDAC, l-ethyl-3-[3-(dimethylamino)propyl]carbodiimide; PhGx, phenylglyoxal; Pi, inorganic phosphate; PPi, inorganic pyrophosphate; PPase, yeast inorganic pyrophosphatase; PRR, proton relaxation rate; CI,CCOOH, trichloroacetic acid; Tris, tris(hydroxymethy1)aminomethane.
0006-296018 110420-6051$01.25f 0 0 1981 American Chemical Society
6052
BIOCHEMISTRY
ions and two phosphates or three Co2+ions and two phosphates to PPase. (4) The structural specificity of site 1 for P, is greater than that of site 2. (5) Several lines of evidence support the suggestion that the binding of a third divalent metal ion is accompanied by a conformational change of the enzyme which has the effect of reducing solvent accessibility to the active site. (6) In enzyme subunits having one or two divalent metal ions bound per subunit, our evidence indicates an absence of divalent metal ion inner sphere binding to PIin either site 1 or site 2 . Experimental Procedures Materials PPase was prepared with some minor modifications (Bond, 1979) of the procedure described previously and had a specific activity of 550-750 IU/mg as measured by the standard titrimetric assay (Cooperman et al., 1973). Trisodium thiophosphate was prepared according to Akerfeldt (1960) and was contaminated with 1-2% P, as estimated by 31PNMR. Methylphosphonic acid was a gift from A. Hampton (Institute for Cancer Research, Fox Chase, PA). Monopotassium phosphoramidate was prepared according to Klement & Becht (1947) as modified by Chambers & Khorana (1958). This material as prepared was free of PI. However, prolonged standing in solution at pH 7 did lead to hydrolysis. For example, samples stored overnight at 4 "C were 2-4% in P,. Phenylglyoxal (PhGx) was obtained from Aldrich Chemical Co. and 1-ethyl-3- [3-(dimethylamino)propyl]carbodiimide (EDAC) from Sigma Chemical Co. Carrier-free [32P]P,was from ICN, and carrier-free 54Mn2+and 57C02+were from New England Nuclear. N-Ethylmorpholine (Pierce, Sequanal Grade) was distilled before use. Dialysis tubing (3787-D52, 12 000 molecular weight cutoff) was obtained from A. H. Thomas and prewashed by boiling with distilled water and rinsing with distilled deionized water. Water for stock solutions was glass distilled and passed through a Barnstead standard mixed-bed deionized column. Stock solutions of Tris (Trizma, Sigma) and KCl used for making up buffers for both the dialysis and NMR experiments were passed over Chelex 100 (Bio-Rad) to reduce trace heavy metal ion contamination. Stock solutions of divalent metal ion salts were prepared by weight from analytical grade samples. Solutions of MnC12.4H20were standardized by using Eriochrome Black T (West, 1969). Methods Equilibrium Dialysis. All dialysis experiments were carried out in buffer A lO.10 M Tris-HC1 (pH 7.2, measured at room temperature) and 0.10 M KCl]. Two methods were employed. Data obtained with method A were used in construction of the Scatchard plots shown in Figures 1 and 2. In this method samples were placed in a Technilab multicavity microdialysis cell (Bel-Art Products) in which each cavity had a volume of 100 pL. In all cases radioactive ligand and enzyme were added to cavities on opposite sides of the dialysis membrane at the beginning of the experiment. The dialysis cells were gently rotated overnight at room temperature (21-23 "C) on a Technilab rotator (Model 71-Bel Art Products), which suitable control experiments showed to be sufficient for the attainment of equilibrium. Following the attainment of equilibrium, aliquots were taken to determine the concentration of radioactive ligand in both cavities and the concentration of enzyme on the enzyme side (this was necessary because the enzyme concentration typically changed during the course of a dialysis experiment). Concentrations of 54Mn2+and 57C02+ were determined bv using a well y counter coupled to a Northern NS
COOPERMAN, PANACKAL, SPRINGS, AND HAMM
600 multichannel analyzer. Concentrations of [32P]Pi were determined by using a liquid scintillation counter as described elsewhere (Springs et al., 1981). Concentrations of PPase were determined by using the protein determination technique of Schaffner & Weissman (1973). Method B was used to prepare samples for 31PNMR measurements, as described previously (Hamm & Cooperman, 1978). Enzyme Inactivation. Rates of PPase inactivation by PhGx or EDAC were determined as previously described (Cooperman & Chiu, 1973b; Bond et al., 1980). All inactivations were carried out in buffer B [0.12 f 0.02 M N-ethylmorpholine acetate (pH 7.0, measured at room temperature)] at 25.0 "C. PhGx concentration varied from 10 to 50 mM. EDAC concentration was always 50 mM. Nuclear Relaxation Rates. The longitudinal proton relaxation rate (PRR) of water ( l / T l ) was measured at 24.3 MHz by a pulsed nuclear magnetic resonance technique described previously (Cohn & Leigh, 1962). Values of the observed enhancements (Eisinger et al., 1962), fob& were calculated from
where 1/ T I and 1/ are the observed relaxation rates in the presence and absence of MnZ+,respectively. The terms with astericks represent the same rates measured in the presence of inorganic pyrophosphatase. In the absence of enzyme, added Pi has essentially no effect on either 1/ T I or 1/ Sample volumes varied from 50 to 100 pL. The longitudinal phosphorus nuclear magnetic relaxation rates of Pi ( l / q ) and of Pi analogues were measured at 40.3 MHz by Fourier transform NMR as described earlier (Hamm & Cooperman, 1978). All NMR experiments were conducted at 25 f 2 "C in buffer A. Enzyme concentration was measured spectrophotometrically by using an extinction coefficient of 1.45 at 280 nm (Kunitz, 1952) for a 0.1% solution of inorganic pyrophosphatase and a subunit weight of 35000 (Cooperman & Chiu, 1973a; Bond et al., 1980). Results We have investigated the binding to PPase of divalent metal ions, Pi, and Pi analogues by four different techniques: equilibrium dialysis, protection of enzyme activity against chemical modification, 31PNMR, and water proton relaxation rates. The results of these studies are presented in turn below. Parameters used in equations are defined either in Table I, in Table VIII, or in the text. Equilibrium Dialysis. The binding of s4Mn2+and s'Co2+ to enzyme in the presence and absence of added Pi was measured by equilibrium dialysis, and the results, plotted in Scatchard form (Scatchard, 1949), are presented in Figure 1. In agreement with the work of Rapport et al. (1 973), we find a stoichiometry of two tightly bound metal ions per subunit for both Mn2+and Co2+. A new result of our work is the demonstration that in the presence of sufficient Pi, three metal ions are tightly bound per subunit. Because of the considerable scatter in most of our data, we have chosen to interpret our results in the simplest possible way, as in each case reflecting binding to intrinsically equal affinity sites. Accordingly, the lines drawn in Figure 1 are linear leastsquares fits to v --- [n
[M2+]
- VI
K
(2)
VOL. 20, NO. 2 1 , 1981
METAL AND P H O S P H A T E B I N D I N G T O P P A S E ~
Table I: Parameter Definitions Conservation and Apparent Equilibrium Constants total concentration in all their [ElT, [Mz+lT, forms of enzyme, divalent metal ion, and Pi or Pi analogue, respectively; specific terms, such as [MnZtIT or [Pi]T are also used total concentration of divalent [M"] B metal ion bound to enzyme apparent dissociation constants for Kp.app9 Kpl,appy Kpz.app Pi or Pi analogue binding to enzyme; subscripts 1 and 2 refer to high- and low-affinity Pi site binding, respectively
Table 11: Apparent Dissociation Constants and Binding Stoichiometries As Determined by Equilibrium Dialysis
Mn'+ Mn'+ Mn'+ coz+ Col+ Pi Pi
k,
0.0103 0.0103 0.00347 0.0165 0.0158 0.0137 0.056
Pi, 0.05 Pi, 4.0 Pi, 2.5 Mnl+, 0.5 CO'+, 1.0
Protection against Inactivation second-order rate constant for inactivation in the absence of added Pi or Pi analogue second-order rate constant for inactivation in the presence of divalent metal ion and saturating Pi or Pi analogue
ko
6053
~~
0.041 0.041 0.0104 0.066 0.048
0.0312 0.143
2.0 2.0 2.9 2.1 3.2 1.05 0.85
60-
31PNMR paramagnetic contribution to Tflobsd T,% T , in the binary Copi complex TP, contribution of enzymebound paramagnetic ion to T% Tpq TI in the complex EMP,, where M is MnZt or CoZt Tpq, T , in the complex EM,P,, where M is MnZ+or Co2+ Water Proton Relaxation Rates enhancement of the EMn complex eb enhancement of the EMnP complex et enhancement of the EMnP, complex eq enhancement per Mn'+ in the EMn,P complex et, enhancement per Mnz+in the EMn,P, complex eqr limit of E o b d at saturating [PIT eS
T%
.
30
c
20
00
'
'
a
0.2 0.4 0.6 0.8 1.0 Y
I
I
A.
2: Scatchard plots of Pi binding to PPase as measured by equilibrium dialysis. (a) In the presence of 0.5 mM Mn2+, enzyme (b) in the presence of 1.0 mM Coz+, concentration 22-36 pM (0); enzyme concentration 21-1 13 p M (0). FIGURE
240 180
im 60
0
0
1.0
2.0
3.0
0
1.0
2.0
3.0
V
FIGURE 1: Scatchard plots of metal ion binding to PPase as measured by equilibrium dialysis. (A) Mn2+ binding: (a) no added Pi, enzyme concentration 7-42 pM (0);(b) in the presence of 50 pM Pi, enzyme concentration 7-36 pM ( 0 ) ;(c) in the presence of 4 mM Pi, enzyme concentration 8-116 p M ( 0 ) . (B) Co2+ binding: (a) no added Pi, enzyme concentration 8-96 pM (0);(b) in the presence of 2.5 mM Pi, enzyme concentration 30-35 p M (0).
where u is the number of divalent metal ions bound per enzyme subunit, n is the number of divalent metal ions bound at saturation, and K is the intrinsic dissociation constant. The apparent dissociation constants for the binding of the first, second, and, where appropriate, third metal ions per subunit, which are related to the intrinsic binding constant by purely statistical corrections, are listed in Table 11, as are the extrapolated values of n. The binding of Pi to enzyme in the presence of Mn2+ and Co2+was also measured by equilibrium dialysis. Scatchard analysis (Figure 2), the results of which are summarized in Table 11, clearly shows the presence of one high-affinity site per subunit. In the absence of divalent metal ions, Pi binding to PPase is weak, with preliminary studies indicating a lower
limit of 0.5 mM for the dissociation constant (data not shown). Protection of Enzymatic Activity against Chemical Modification. We have previously shown that PPase is inactivated by PhGx and that although Mn2+ or Ca2+ alone does not protect against such inactivation, the combination of either Mn2+ and the competitive inhibitor hydroxymethanediphosphonate or Ca2+and PPI afford strong protection (Cooperman & Chiu, 1973b; Bond et al., 1980). This approach has now been extended to examine whether Pi and methylphosphonate protect the enzyme from PhGx inactivation. Under pseudo-first-order conditions (excess PhGx), secondorder rate constants (kOw)were determined for PhGx inactivation of PPase at fixed metal ion concentration and varying Pi or methylphosphonate concentrations. The results obtained, when plotted in double-reciprocal fashion according to eq 3 (Figure 3), yield apparent dissociation constants, Kp,app,for Pi and methylphosphonate 1 1 Kwpp - -1 --+-(3) 1-a 1-b 1 - b [PI where [PI is equal to the Pi or methylphosphonate concentration, a is equal to kOw/ko,b is equal to k,/ko, and ko and k, are the second-order rate constants for inactivation of the enzyme in the absence of Pi or methylphosphonate and in the presence of saturating Pi or methylphosphonate, respectively?
* We interpret the dependence of the rate of PhGx inactivation of enzyme as a function of Pi concentration as reflecting a single Pi binding process (Figure 3). This procedure is necessary because our data are insufficiently precise to justify an attempt to define two binding processes but will yield a Kpcppintermediate between Kplcppand Kpzcppif the binding of the first Pi confers only partial protection.
B I ocH E M I sT R Y
6054
COOPERMAN, PANACKAL, SPRINGS, AND HAMM
c
I
so 0.3
0.2 1
1
1
1
l
l
1
1
1
l
1
1
1
1
80 120 160 200 240 la1 4 6 8 10 12 i b c ) 0 2 0 4 0 6 0 8 1 0 1 2 !d e ) 40
2
1
PI],
!mM
0.1
-
-
0
I
I
4
7l
75!i
6
150
1
0 0.0 02
04 06 08
10 12 14 16
18 20
22
LEI,
Table 111: Studies on Protection against Chemical Inactivation
[E]T 0.0023 0.028 0.028 0.028 0.028 0.027 0.033 0.0095 0.025 0.030
[fixed ligand] (mM) Mn2+,1.0 CO'+, 1.0 M$:, 10 Zn , 1.0 Mn2+,1.0 Pi, 3.0 Mni+, 0.4 Co2+,1.0
ko
ks
(M-' (M-' Varying ligand reagent min-') min-')
Pi Pi Pi Pi CPR
Pi MgZ+,10 Pi Zn2+,0.9 P,
PhGx PhCx PhGx PhGx PhGx PhGx EDAC EDAC EDAC EDAC
1
2 0.2
1
1
1
1
1
1
1
1
4 6 8 10 0.4 0.6 0.8 1.0 l i I P i 1 , (rnM-')
1
1
1
12 ( a , b ) 1.2 IC)
Rates of EDAC inactivation of PPase, plotted according to eq 3. (a) In the presence of Coz+and varying Pi (A); (b) in the (c) in the presence of Mgz+ presence of Zn2+ and varying Pi (0); and varying Pi (0). FIGURE 6:
4: Rate of PhGx inactivation of PPase as a function of [P,]T/[E]T. [E]T, 0.37 mM; [Mn2+]~,4.0 mM.
FIGURE
(mM)
I
t
FIGURE 3: Rates Qf phenylglyoxal inactivation of PPase, plotted according to eq 3. (a) In the presence of Mn2+ and varying PI (0); (b) in the presence of Coz+and varying Pi (0); (c) in the presence of Zn2+and varying Pi (A); (d) in the presence of Mg2+and varying PI ( 0 ) ;(e) in the prdsence of Mn2+and varying methylphosphonate (A).
0
1
1.15 1.28 1.29 1.08 1.15 0.60 0.92 2.54 2.58 1.78
KP*aPP
(mM)
0.21 0.23 0.31 0.31 0.68
0.0094 0.090 2.4 0.111 4.1
0.68 0.20 0.35
0.058 4.5 9.18
___ ..
Metkylphosphonate.
Kp,Bpp, ko, and k, values are listed in Table 111. As expected, ko values are almost independent of the identity of added divalent metal ions, reflecting the lack of protection qffered by metal ions alone. By contrast, added P, alone (3 mM) affords significant protection, providing evidence that PIbinds to enzyme even in the absence of added metal ion. In the presence of Mn2+ protection by PI is exerted at especially low PI concentration, and, accordingly, the Mn2+-P, system was investigated further. A plot of kOMvs. [PJT/[ElT at fixed [Mn2+ITand an enzyme concentration (0.37 mM in subunits) large compared to Kp,appis presented jp Figure 4. The stoichiometry of PI binding per subunit necessary for protecfion can be estimated as being equal to 1.01, measured
as the point of intersection of the initial slope and the line corresponding to k, (see Table 111). The rate of PhGx inactivation of PPase was also studied at fixed Pi concentration and varying Mn2+/enzyme ratios, as shown in Figure 5. The results clearly indicate that full protection requires, in addition to Pi, more than one bound Mn2+/enzyme. Below we will show the data prqented under Results permit estimation of all of the relevant constants describing Mn2+ and Pi binding to enzyme (see Table VIII). These estimates, in turn, permit calculation of the concentration of various enzyme species at given [E],, [Mn2+IT,and [Pi]T. In the presence of 3 mM Pi, the dominant enzyme species in solution during the course of the titration with Mn2+ are EPi, EMnPi, EMn2(Pi)p,and EMn3(PJ2. The initial and plateau values of kow provide estimates of the second-order rate constants for EPi and EMn3(Pi),, which are 0.60 and 0.12 M-* min-', respectively. A simple curve-fitting procedure allows estimation of rate constants of 0.60 and 0.38 M-' min-' for EMnPi and E M I I ~ ( R ~respectively. )~, The line in Figure 5 is drawn by using these estimates. These results show that maximal protection requires, in addition to the binding of Pi, the binding of three Mn2+ per subunit. Second-order rate constants for enzyme inactivation were also obtained in the presence of excess EDAC, fix$ metal ion concentration, and varying Pi concentration (Figure 6) and analyzed as described above for the PhGx studies, with the results summarized in Table 111.
V O L . 20, NO. 21, 1981
M E T A L AND P H O S P H A T E B I N D I N G TO P P A S E Table IV:
6055
Longitudinal Rate Studies [E]T (mM)
[fixed ligand] (mM)
0.94 1.00 0.10 0.10
Co2+,0.05 Cof+,0.50 Co'+, 1.8 Mn2+,0.040 Mn'+, 0.040
0.10 This work (Figure 7).
variable ligand
Kp,app (mM)
1/T%o!nPlex (s-')
ref
4.6 11.5 3.8 136 6.7
1/Tlb, 3.6 x 10' l/Tlq, 1.1 X 1.0' l/Tlqt, 2.7 X 10' i / ~5.1~ x ~103,
a a a
35
i p I q , 2.3 x 103 i / ~4.4~x 103 ~ ,
Pi Pi Pi SPC
CPd
Mn'+, 0.015
Pi
Hamm & Cooperman, 1978.
18
.
Thiophosphate.
a
a a
b
Methylphosphonate.
Table V: Evaluation of TFc by the Dialysis Method fE1T (mM) 1.0 1.0
[CO'+IT(mM) 1.8 1.8
[Pi]" (mM) 5.0 50
Tfobsd (enzyme side) Ttobsd (nonenzyme side) 0.033 f 0.003 1.5 f 0.25 0.20 f 0.020 1.8 i7 0.1
TPC 0.034 0.232
31PLongitudinal Relaxation Rates. We previously dowed that Cp,the paramagnetic contribution to q , o b s d , could be used to measure the interaction of EMn with Pi in solutions of enzyme, Mn2+,and Pi (Hamm & Coopermad, 1978). This approach is now extended to study Pi interactions with ECo. We first determined the relaxation time q b ahd the dissociation constant KB for the binary CoPi complex. $Atconstant [coz+]T(0.050 mM), Tpp varied linearly as a function of added Pi (10-40 mM), permitting calculation of TpPband KB according to eq 4 (Navon et al., 1970) (see Table IV) [ c o 2 ' l T ~ p = qb(KB + [pi])
(4)
In our previous work we found that the binding of Pi to the high-affinity site (site 1) of EMn made no apparent contribution tosTpp so that such measurements could be used to calculate Kwm and qq according to eq 5, where K, reflects [MZ'lBTPc = TPq(Kp2,app + [PI)
(5)
binding to the low-affinity site (site 2), Tp, is the bontribution of enzyme-bound paramagnetic ion to Tpp, Tpp is the relaxation time in the complex EMP,, and [MZfIBis the total concentration of divalent metal ion bound to enzyme. As seen in Figure 7, Tp, is also a linear function of added Pi in solutions of enzyme and Coz+. The slope and intercept of line a allow calculation of Kh,appand qqfor ECoP,, and these values in from the slope and turn permit calculation of Km,appand qq, is the relaxation time in the intercept of line b, where qqt complex ECo2P2. These values are collected in Table IV. qpvalues have also been measured to study Pi analogue binding to EMn. The linear behavior seen as a function of thiophosphate concentration (Figure 7,line c) allows calculatiou of Kpa, and qqvalues for this analogue. qqis similar to what was found previously for Pi, which may be taken as evidence that thiophosphate binds to site 2 with the same orientation relative to Mn2+ as does PI, bul the affinity of thiophosphate for the enzyme is -9-fold less compared to that of Pi (Table IV). In the case of methylphosphonate, a more complex curve is observed (Figure 7,link d). A simple model consistent with the observed concentration dependence of qp involves obligatory ordered binding of' methylphosphonate to site 1 prior to binding to site 2, with only site 2 binding contributing to Tp,. This model is described by
where Kplappand K,,app are the dissociation constants for high a f f i t y site and low affinity site bind8hg,respectively. It should be noted that if Kpl,app [MnZ+IT,allowed measurement of two dissociation constants, corresponding to PI binding to Sites 1 and 2, as well as of enhancement values for the ternaty EMnP and quaternary EMnP2 complexes (Hamm & Cooperman, 1978). In the present work we first extend this approach to examine the interactions of bhosphate analogues with EMn. The results are presented in Figure 8.
B I oc H E M I S T R Y
6056
COOPERMAN, PANACKAL, SPRINGS, A N D H A M M
3.0 3
I PI or PI analogue]
1
I
I
I
I
I
I
0.10
0.040
Pi
0.10
0.040 0.040 0.040 0.040 0.20
CPc
0.10 0.10
1 2 -
0.10
CP t 2 mMPi
SPd NPe
0.24 14 32 4.3 853 >72
0.95
Pi
*
Hamm & Cooperman, 1978. This work. nate. Thiophosphate. e Phosphoramidate.
0
0 12 12
0 1 6 (a-ci 16 (d)
ImM ' I
FIGURE 9: Dependence of cow on Pi or P, analogue concentration, plotted according to eq 7. Lines a-d are as described in the legend to Figure 8. Lines drawn are linear least-squares fits to the results.
For all three analogues, only a rather low affinity binding mode is apparent. The simple phenomenological eq 7 adequately 1 t b - cobsd
-
l tb
-
+ Es
KpmJ ( e b - %)[PI
7.7 (et) 5.5 (eq) 5.1 2.1 5.1
4.4 5.5 (et,) 3.45 ( E , , )
a a
b b b b b b
Methylphospho-
the complex EMnP (equal to 7.7;Hamm & Cooperman, 1978), replaces t b in eq 7. The variation of with Pi was also studied at 2:l [Mn2']/[E]~ (Figure 10). When [Pi]T 2 1.0 mM, essentially all the enzyme in solution contains at least one bound Pi and eq 8 is approximately valid, where et' is the average en-
I
008 08 1 ,P
I _
Table VI: Water Proton Relaxation Rates
0.10
004 04
I
(mu)
FIGURE 8: Dependence of cowon Pi or Pi analogue concentration. All experimentscontained 0.10 mM enzyme and 0.04 mM Mn2+. (a) Varying thiophosphate (0);(b) varying phosphoramidate (A); (c) varying methylphosphonate(0); (d) varying methylphosphonatein the presence of 2 mM Pi ( 0 ) ;(e) varying Pi [redrawn from Hamm & Cooperman (1978)l (0).
0
I
(7)
describes the concentration dependence of tow, where t b is the enhancement of the EMn complex and e, is the limit of e o b d at saturating [PIT. The reciprocal in Figure 9 allow evaluation of Kpppqand 6,. These values are collected in Table IV. For both thiophosphate and phosphoramidate Kp,appvalues are lower limits, since even minor (1-2%) contamination with Pi would give curves similar to those obtained even if neither thiophosphate nor phosphoramidate had any affinity for enzyme-Mn2+. For methylphosphonate, there is no Pi contamvalue of 32 mM is reliable. This value ination, and the Kp,app measured in the 3'P studies decorresponds closely to Kp,app scribed above (Table IV) and presumably reflects binding to site 2. Thus, in contrast with Pi, methylphosphonate binding to site 1 results in little apparent change in t&d. As can be seen by comparison of curves c and d in Figure 8, the addition of 2 mM Pi considerably enhances EMn affinity for methylphosphonate. We interpret this result as showing that methylphosphonate binds much more strongly to site 2 when Pi replaces methylphosphonate in site 1. Accordingly, for the reciprocal plot d in Figure 9, et, the enhancement value for
-- 1 tobsd
- cq'
- ++1) et'
-
(8) Kp,app
hancement per Mn2+ in the complex EMn2P and tql is the average enhancement per Mn2+in the complex EMn2P2. The value of tg' equal to 3.45was estimated directly from the results shown in Figure 10 as the limiting value of tow at saturating [Pi]. A plot of [l/(tow - e;)] vs. [Pi] gave a straight line (Figure 10, inset) permitting evaluation of KPBmand e{. These values are listed in Table VI. Finally, measurements of tobsd when [Mn2+IT> [E], and [E], is large compared to the dissociation constant for Pi binding determined by equilibrium dialysis (Table 11) permitted estimation of the stoichiometry of high-affinity Pi binding. This estimate is made from the point of intersection of a line drawn with the initial slope of the tow vs. [PilTcurve and a line of zero slope drawn as the limiting value of tow at saturating inorganic concentration (Figure 1 1) as determined by a small1 extrapolation of the data to infinite [PIT. For curve a (0.060 mM subunit), the limiting value of cow is 1.47 and the point of intersection occurs of 0.060 mM inorganic phosphate, giving a site stoichiometry of 1.OO/subunit. For curve b (0.10mM subunit), the limiting value of tow is 1.76 and the point of intersection occurs at 0.085 mM inorganic phosphate, giving a site stoichiometry of 0.85/subunit. Interpretation of Results and Discussion This section is divided into four parts. In the first we show how the equilibrium dialysis and 31PNMR data presented in
VOL. 20, NO. 21, 1981
M E T A L A N D P H O S P H A T E B I N D I N G TO P P A S E
001
0.05
009
[PI,
013
021
017
(mM1
6057
KpZM,= A I P i l b - lpila 1-A
11: Dependence of €"%on [Pi], at fixed [Mn], and [E],. (a) [ElT,0.060 mM, and [Mn IT, 0.60 mM; (b) [E],, 0.10 mM, and [Mn2+IT,0.60 mM. FIGURE
Scheme I: Relevant Equilibria in Solutions of Enzyme, Divalent Metal Ion, and Pi E
Kh e EM
-
EP
5 EMP
e EMzP e EM3P
KM2
d
EM2
(EM3)
Table VIII: Dissociation Constants values (mM) definitions
none
MnZ+
Co"
0.80
this and an earlier paper (Hamm & Cooperman, 1978) allow us to evaluate all of the important equilibrium constants describing Mn2+, Co2+, and Pi binding to PPase and compare KwPpvalues predicted by these evaluated constants with those measured by protection of enzymatic activity against inactivation and by a PRR experiment. In the second, we consider the evidence for a conformational change on the enzyme accompanying the binding of a third metal ion. In the third we discuss phosphate and phosphate analogue binding to enzyme and the variation in affinity with the identity of added divalent metal ion. Finally, we conclude with a discussion of what we presently know about the relative locations of Pi and divalent metal ions on the enzyme surface. Evaluation of Equilibrium Constants. The equilibrium dialysis results presented in Figures 1 and 2 and Table I1 and the stoichiometric titrations seen in Figures 4 and 11 show clearly the presence of one high-affinity Pi site per enzyme subunit (site 1) and also show that, in addition to the two high-affinity divalent metal ions bound per PPase subunit previously found by Rapoport et al. (1973), a third divalent metal ion binds in the presence of added Pi. The overall enzymatic reaction requires an additional Pi site (site 2), and direct evidence for such a site has been presented by us previously (Hamm & Cooperman, 1978; Springs et al., 1981). Therefore, in solutions of enzyme, Mn2+or Co2+,and Pi we have potentially 12 species to consider, as shown in Scheme I. In fact, in our treatment we consider EM3 to be stoichiometrically insignificant over the concentration ranges we employ, on the basis of the Scatchard plots for Mn2+and Co2+ binding in the absence of Pi (Figure 1). EP2 is also considered negligible, because of the strong dependence on metal ion of Pi binding to site 2 (see Table VIII), the already weak affinity of Pi for EMP, and the relatively low Pi concentrations employed in the dialysis studies. Finally, we note that the con-
20 0.0103 0.041 0.24 18 0.25 0.75 0.029 0.66
4.6 0.0165 0.066 0.54 11.5 0.32 3.8 0.21 0.82
centration term [EM3P2Itrepresents the sum of [EM3P2]and [EM3PPi]. Elsewhere, we have shown, at least for Mg2+,that appreciableamounts of EM3PPiare formed in equilibrium with EM3P2(Springs et al., 1981), and the studies presented in this paper do not distinguish between these species. There are a total of nine independent equilibrium constants which link the remaining 10 species. For Mn2+,two of these, KpMand KpzM,have been determined by us previously by water proton relaxation rate measurements (Hamm & Cooperman, 1978), and two others, KM, and KMz,may be estimated directly from the Scatchard plot of equilibrium dialysis results for MnZ+ binding in the absence of Pi (Table 11). Further equilibrium dialysis measurements of Mn2+binding to enzyme in the presence of Pi and of Pi binding to enzyme in the ~, K M ~ ~ ~ ~ , presence of Mn2+provide values for K M , , ~ ?KM2,app, and KP,app,(Table 11) permitting estimation of the five reKMMZP,and maining equilibrium constants Kp, KpMz,KpZM2, KPzM3from eq 9 and 11-14 (Table VII). Estimates for all nine constants are presented in Table VIII. For Co2+, KM, and KM2 are estimated directly from the Scatchard plot of equilibrium dialysis results for Co2+binding in the absence of Pi (Table II), and Kp2Mand Kp2%are estimated directly from the concentration dependence of the 31P longitudinal relaxation rates (Table IV). With Kp already estimated from the results with Mn2+, the remaining four
6058
B I ocH E M I s TR Y
COOPERMAN, PANACKAL, SPRINGS, AND H A M M
I____
Table IX: Comparison of Predicted and Experimental K,.,,,
(mM) Values predicted
expt
1 2 3 4 5 6
no. -
method PhGx protection PhGx protection EDAC protection water PRR PhCx protection EDAC protection
________ iF1T (d)_________ [ M z f I T (mM) 0.0023 0.028 0.0095 0.10 0.028 0.025
MnZ+,1.0 C o l t , 1.0 coz+, 1.0 Mn2+,0.20 Me2+,10 Mg*+. 10
constants, KpM,KpM2,KMM2', and Kp2M3,can be estimated from ~ ~Kplaps ~ ~values (Table 11) measured KMl,app, KMzapp, K M , and and eq 10 and 12-14 (note that eq 10 IS just a rearranged version of eq 9). Estimates for these constants are also presented in Table VIII. The values in Table VI11 enable us to predict Kappvalues for Pi sites 1 and 2 at given values of Pi and Mn2+ or Co2+ concentrations. In Table IX such predicted values are compared with KPrqpvalues measured under conditions similar to (for the protection against inactivation of enzymatic activity experiments) or identical with (for the PRR experiment) those used in determining the values given in Table VIII. Also and Kp2Fpp values determined compared in Table IX are ICPlaPP in the presence of Mg2+from an analysis of the dependence of enzyme-bound PPI formation on Pi concentration (Springs et al., 1981) with KP,appvalues measured in the PhGx and EDAC protection experiments. The Kappfrom the PRR experiment, performed in the presence of Mn2+,agrees fairly well with the predicted value for KP2,app.This is an important consistency check for our analysis, since it represents a direct determination of what is essentially KplMlin the presence of Mn2+,a constant which we had otherwise evaluated only indirectly from eq 11. For the PhGx protection experiments in the presence of Mn2+or Co2+, the closeness of agreement between KP,appand the values provides strong evidence that it is P, predicted KP,,app binding to site 1 alone that results in protection of an essential arginine (Bond et al., 1980) which we elsewhere have shown to be Arg-77. Given the known ability of arginine to hydrogen bond to P, (Cotton et al., 1974), it would be reasonable to attribute the observed protection to the direct interaction of Pi in site 1 with Arg-77. Evidence for a Conformational Change. Although Arg77-Pi (site 1) hydrogen bonding may indeed by necessary for protection, other results lead us to suggest that maximal protection requires, in addition, a conformational change accompanying the binding of a third divalent metal ion, which makes the enzyme active site less accessible to solvent. The evidence for this suggestion is 3-fold. First, there is the direct experiment presented in Figure 5, showing that although P, binds to PPase in the absence of added divalent metal ion and confers partial protection, maximal protection requires three enzyme-bound Mn2+. Second, the value of KP,appmeasured in the presence of Mg2+ by protection against PhGx inactivalues found in vation falls in between the KPl,appand Kp2,app measurements of enzyme-bound PPi formation, even allowing for small differences which could arise from differences in the reaction media for the two experiments. We interpret these results as showing that, in the presence of Mg2+,the binding of a single Pi affords only partial protection, with full protection requiring the binding of both phosphates per subunit. This is consistent with our suggestion, since unlike the binding of a third Co2+or Mn2+,which requires the binding of only a single phosphate per subunit, the binding of a third MgZCto the enzyme (at 10 mM [Mg2+])requires the binding of two
K~,,app
0.0070 0.056 0.056 0.24 0.5 2 0.5 2
KPz,aPP
0.66 0.82 0.82 1.59 21 21
measured K,,,p 0.0094 0.090 0.058 0.95 2.4 4.5
phosphates per subunit (Springs et al., 1981). Third, the average value per Mn2+ of the PRR enhancement parameter progressively decreases for the complexes EMnPi, EMnzPi, and EMn3Pi, having the value 7.7 for EMnPi (Hamm & Cooperman, 1978), 5.5 for EMn2Pi(Table V), and 2.4 for EMn3Pi (as estimated from the plateau values in Figure 11). This trend supports our suggestion, since t provides a measure of the accessibility of solvent water to enzyme-bound Mn2+. It might be argued that the fact that KP,appvalues measured by protection against EDAC inactivation are similar to or identical with those measured by protection against PhGx inactivation provides a fourth piece of evidence for a conformational change restricting solvent access to the active site, since charge repulsion should prevent the carboxylate side chains that are the most likely targets of EDAC modification from interacting directly with Pi. However, this result could also reflect direct metal ion interaction with these side chains, since, under the conditions of our experiments, the binding of the third divalent metal ion occurs concurrently with phosphate binding (Figure 1).
Phosphate and Phosphate Analogue Binding. Phosphate affinity for site 1 varies considerably as a function of the identity of added divalent metal ion, a result which we can, as a result of our analysis of divalent metal ion and Pi binding, attribute directly to differences in the binding of the third divalent metal ion and more specifically to differences in the value of KMM2P. Thus KPl,appis 0.009-0.014 mM in the presence of Mn2+(0.5-1.0 mM), 0.06-0.09 mM in the presence of Co2+ (1 .O mM), 0.1 1 mM in the presence of ZnZ+ (assuming that, as with Mn2+ and Coz+, Pi binding in site 1 protects against PhGx inactivation), and 0.5 mM in the presence of 10 mM Mg2+,and KMMzPfollows the same trend, 0.029 mM for Mn2+,0.22 mM for Coz+(Table VIII), and >10 mM for Mg2+(Springs et al., 1981). By contrast, the values of KPMfor Mn2+ and Coz+ and of KpM2for Mn*+, Coz+, or Mg2+(same as Kplsppat 10 mM Mg2+) differ little from each other or from Kp (Table VIII). Phosphate affinity for site 2 also varies as a function of divalent metal ion, being considerably less in the presence of Mgz+ than in the presence of either Co2+or Mn2+ (Table IX). The effect of divalent metal ion on relative Pi affinity thus shows an inverse correlation with the effectiveness of divalent metal ion as a cofactor for catalytic enzyme activity, which for both PPi hydrolysis and water-Pi oxygen exchange falls in the order Mg2+ > Zn2+ > Coz+ Mn2+(Janson et al., 1979; Butler & Sperow, 1977). This is perhaps not surprising since higher affinity is consistent with a slower dissociation rate, and elsewhere we (Springs et al., 1981) and others (Hackney & Boyer 1978; Cohn & Hu, 1978; Hackney, 1980) have shown that release of each Pi from enzyme is partially rate determining in overall enzyme-catalyzed PPi hydrolysis. Studies of Pi analogue binding to PPase were undertaken in order to provide information on the structural requirements of sites 1 and 2. PRR and 31Pstudies show that both phosphoramidate and thiophosphate have lower affinity for PPase
V O L . 20, N O . 2 1 , 1981
METAL A N D P H O S P H A T E B I N D I N G TO P P A S E
than Pi. Since interpretation of experiments with these analogues would be complicated by the presence of even minor contaminating amounts of Pi, we have concentrated our efforts on methylphosphonate, where such contamination is not a problem. Our results indicate that methylphosphonate is a much poorer ligand than Pi for site 1 but is comparable as a ligand for site 2. Thus, when [Mn2+ITC [ElT, the apparent dissocation constants for site 1 are 0.24 mM for Pi and 6.7 mM for methylphosphonate (Tables IV and VI). Similarly, when [Mn2+IT> [ElT, the apparent site 1 dissociation constants measured by protection against PhGx inactivation are 0.0097 mM for Pi and 4.1 mM for methylphosphonate (Table 111). This large difference in affinity may be due at least in part to an inability of the enzyme to undergo the putative conformational change reducing active site accessibility to solvent when methylphosphonate replaces Pi. This possibility is suggested by the result that the protection against PhGx inactivation afforded by saturating methylphosphonate is much less than that afforded by saturating Pi. On the other hand, the apparent dissociation constants for site 2 when [Mn2+IT C [ElTare similar: 14-18 mM for Pi and 32-35 mM for methylphosphonate (Table IV and VI). Moreover, in the presence of 2 mM Pi, which is sufficient to almost saturate site 1 while site 2 remains essentially open, the apparent dissociation constant for methylphosphonate, which should measure binding to the low-affinity site, is 4.3 mM, that is, considerably lower than the value of 14-18 mM for Pi itself. Metal Ion to Pi Distances on PPase. In this and previous work (Hamm & Cooperman, 1978; Springs et al., 1981), we have demonstrated the presence, per PPase subunit, of two sites for Pi binding of different affinities and specificities and shown that the high-affinity site 1 contains an essential arginine. We also have shown that three divalent metal ions bind per enzyme subunit, and are necessary for enzymatic activity, and presented evidence for a conformational change accompanying the binding of a third divalent metal ion. It now remains to ascertain the functional roles of these metal ions in the catalytic process. A first goal, which has thus far only been partially achieved, is the determination of the distances between the bound divalent metal ions and Pi in sites 1 and 2, and it is with a discussion of our current knowledge on this point that we will conclude this paper. Previously, we used the measured effect of enzyme containing one Mn2+ per subunit on the Tl of enzyme-bound Pi to calculate a Mn2+I3lPdistance of 6.2 f 0.4 A, consistent with a second sphere complex between the enzyme-bound Mn2+and Pi in site 2 (Hamm & Cooperman, 1978). In view of our current results showing that in the presence of Pi, the intrinsic binding constants for all three Mn2+ which can be bound are roughly equivalent, this distance is actually a weighted root mean sixth average for metal ion bound in all three potential sites. This raises the possibility that one divalent metal ion site, occupied only partially, is responsible for all of the observed relaxation and thus is considerably closer to Pi than 6.2 A. On the basis of the near equivalency of divalent metal ion site affinities, it is reasonable to assume that no one of the three sites contains 7 ms. In the present study, measurements of the effect of enzyme-bound Co2+on the qpwere used principally to estimate Kp2Mand KPIMzand are incomplete with respect to calculation of Co2+-Pi distances. Nevertheless, they do allow some conclusions to be drawn. First, we see no evidence for Co2+ interaction with Pi in site 1 in the complex ECoP, which would have shown up as a negative deviation from line a in Figure 7 at low Pi concentration (line b, at a Co2+/enzyme subunit ratio of 1.8:1.0, was not extended to low enough concentration to test for such a deviation). We estimate that a mean residence time of Pi in site 1 > 50 ms would be required for this lack of observed effect to result from slow chemical exchange. Given the value for KPMof 0.54 mM (Table VIII), this in turn would require a second-order binding constant of Pi to ECo of 3 x 104 M-' s-1 , which is extremely slow for simple Pi binding to a protein. This result strengthens the argument that in an enzyme subunit containing one bound divalent metal ion, the divalent metal ion is not close to site 1 . Second, the qq and qq. values measured respectively for ECoP2 and ECo2P2 (Table V) allow calculation of lower limits for the root mean sixth Co2+-Pi (site 2) distance, r, on the basis of the assumption that of the four mechanisms which can contribute to 31Prelaxation [diamagnetic relaxation, outer sphere relaxation, dipolar interaction, and hyperfine contact (Mildvan & Gupta, 1979)] only dipolar interaction is important. Thus r is given by inequality 15, where wIand usare the nuclear and electronic
precession frequencies, respectively, T , is the correlation time for the dipolar interaction, and 658 f 132 is the numerical value of a constant term for C O * + - ~ ~interaction P (Mildvan & Gupta, 1979). The Co2+ complexes of enzyme are pink, characteristic of high-spin Co2+. For example, for the ECo2P complex the relevant spectral parameters are as follows: A, (nm) 460,492, 550 (6 14, 14,8). The electron spin relaxation time of high-spin Co2+,which should dominate T,, ranges from 5 X to IO-" s. The value of qqfor ECoP2, 0.0088 s (Table V), thus gives a value for r 1 3.9-5.4 A, where 1 3 . 9 A corresponds to a T~ of 5 X sand 1 5 . 4 A to a T~ of 10-" s. The value of qqfor ECo2P2is 0.0038 s (Table V). If both Co2+ions are assumed to interact equally with Pi in site 2, then from inequality 15, r 2 3.8-5.2 A; if only one Co2+interacts, then r 1 3.4-4.7 A. As these estimates make clear, it will be of obvious value to extend our measurements to the ECo3P2 complex and to evaluate both T , and the magnitude of the dipolar contribution to 31Prelaxation in order to calculate r more accurately. Such measurements are currently under way. Acknowledgments We acknowledge the assistance of Ben Kruskal, who helped
6060
Biochemistry 1981, 20, 6060-6065
with some of the experiments described above.
Hackney, D. D., & Boyer, P. D. (1978) Proc. Natl. Acad. Sci. U.S.A. 75, 3133-3137. Hamm, D. J., & Cooperman, B. S. (1978) Biochemistry 17, 4033-4040. Janson, C. A., Degani, C., & Boyer, P. D. (1979) J . Biol. Chem. 254, 3743-3749. Klement, R., & Becht, K. H. (1947) 2. Anorg. Chem. 254, 2 17-220. Kunitz, M. (1952) J . Gen. Physiol. 35, 423-450. Lietzke, M. H. (1963) A Generalized Least-Squares Program for the IBM 7090 Computer, DRNL-3295, Office of Technical Service, U S . Department of Commerce, Washington, DC. Mildvan, A. S., & Grisham, C. M. (1974) Struct. Bonding (Berlin) 20, 1-21. Mildvan, A. S., & Gupta, R. K. (1979) Methods Enzymol. 49, 322-359. Navon, G., Shulman, R. G., Wyluda, B. G., & Yamane, T. (1970) J . Mol. Biol. 51, 15-30. Rapoport, T. A., Hohne, W. E., Heitmann, P., & Rapoport, S . (1973) Eur. J . Biochem. 33, 341-347. Scatchard, G. (1949) Ann. N . Y . Acad. Sei. 51, 660-672. Schaffner, W., & Weissmann, C. (1973) Anal. Biochem. 56, 502-514. Springs, B., Welsh, K. M., & Cooperman, B. S. (1981) Biochemistry (in press). West, T. S. (1969) Complexometry, B. D. H. Chemical Ltd., Poole, England.
References Akerfeldt, S . (1960) Acta Chem. Scand. 14, 1980-1984. Bond, M. W. (1979) Ph.D. Thesis, University of Pennsylvania, Philadelphia, PA. Bond, M. W., Chiu, N. Y., & Cooperman, B. S. (1980) Biochemistry 19, 94-102. Butler, L. G., & Sperow, J. W. (1977) Bioinorg. Chem. 7 , 141-150. Chambers, R. W., & Khorana, H. G. (1958) J . Am. Chem. SOC.80, 3749-3752. Cohn, M., & Leigh, J. S . (1962) Nature (London) 193, 1037-1 040. Cohn, M., & Hu, A. (1978) Proc. Natl. Acad. Sci. U.S.A.75, 200-203. Cooperman, B. S., & Chiu, N. Y . (1973a) Biochemistry 12, 1670- 167 5. Cooperman, B. S., & Chiu, N. Y. (1973b) Biochemistry 12, 1676-1682. Cooperman, B. S., Chiu, N.Y., Bruckmann, R. H., Bunick, G. J., & McKenna, G. P. (1973) Biochemistry 12, 1665-1669. Cotton, F. A,, Day, V. W., Hazen, E. E., Jr., Larsen, S., & Wong, S. T. K. (1974) J . Am. Chem. SOC.96,4471-4478. Eisinger, J., Shulman, R. G., & Szymanski, B. M. (1962) J. Chem. Phys. 36, 1721-1729. Hackney, D. D. (1980) J . Biol. Chem. 255, 5320-5328.
Use of Phospholipase D To Alter the Surface Charge of Membranes and Its Effect on the Enzymatic Activity of D-P-Hydroxybutyrate Dehydrogenase+ Robert M. Clancy, Alan R. Wissenberg, and Michael Glaser*
ABSTRACT:
The effect of an electrostatic potential on the enzymatic activity of D-P-hydroxybutyrate dehydrogenase was examined. Phospholipase D was used to increase the surface charge and concomitantly the electrostatic potential of submitochondrial membranes. The apparent K, for the negatively charged substrates of D-P-hydroxybutyrate dehydrogenase increased as the membranes were reacted with phospholipase D. There was a 10-fold increase in the apparent K , for NADH when the content of acidic phospholipids was increased by 24%. The addition of monovalent or divalent cations, which reduced the electrostatic potential, largely reversed the apparent K,,, changes. At the same ionic strength, divalent
cations had a substantially larger effect than monovalent cations. Similar results were obtained when the purified apoenzyme was reconstituted in unilamellar vesicles containing different ratios of phosphatidylcholine and acidic phospholipids. When the apoenzme was reconstituted into phosphatidylcholine vesicles containing increasing amounts of phosphatidylethanolamine, the apparent K, also increased but to a smaller extent, and increasing the ionic strength did not reverse this effect. The results show that the apparent K , of D-phydroxybutyrate dehydrogenase can be significantly altered by an electrostatic potential as well as other properties of the phospholipid polar head group.
Electrostatic potentials have been shown to affect the catalytic activity of enzymes in a variety of systems (Douzou & Maurel, 1977). One approach to studying these effects has been to immobilize trypsin and chymotrypsin to an insoluble polyanionic carrier (Goldstein et al., 1964; Goldstein, 1972). The polyanionic environment was found to promote a strong
local negative electrostatic potential which acted on the chemical properties of the medium. In particular, the local concentration of ionic species were found to be different from those of the bulk solution. Another study employing lysozyme in the polyanionic environment of the cell wall of Micrococcus luteus has shown similar results (Maurel & Douzou, 1976). Relatively few studies have been carried out on the effect of an electrostatic potential on the properties of membrane-bound enzymes. Recently, Wojtczak & Nalecz (1979) and Nalecz et al. (1980) used detergent methods and fatty acids to increase the electrostatic potential of a biological membrane and observed the effect on several membrane-bound enzymes.
From the Department of Biochemistry, University of Illinois, Urbana, Illinois 61801. Received April 16, 1982. This work was supported by National Institutes of Health Grant GM 21953. M.G. was supported by National Institutes of Health Research Career Development Award GM 00193.
0006-296018 110420-6060$01.25/0
0 198 1 American Chemical Society